The Nineteenth Byte

att

12:28 AM

the animation led me to expect to be able to click on the language list buttons/icons

TKirishima

I think that I canceled them. Maybe if you hard reload they won't be there anymore

Adám

@TKirishima APL's theme colour would be the green logo's green.

TKirishima

Ok!

Adám

And J's is a turquoise, see jsoftware.com

TKirishima

Thank's ! I'm going to fix that

Zion mycelia adamancy

12:47 AM

@TKirishima Looks cool! I speak more about it here, but I personally dislike golfing sites with a set list of a languages. In short, I feel like it hampers both golfing creativity, and the potential userbase of the site (more languages = more users). I'm not sure how you do the language selection/backend stuff, but if possible, I'd open it to a lot more languages

Adám

The plan is adding a language per week, so all TIO languages will be supported in about 13 years or so ;-)

Seggan

so no newer langs?

Adám

How many newer languages are there? Enough to make it 14 years?

Zion mycelia adamancy

Also, a few aesthetic things:
when I scroll down on `/index.php`, I get a weird jump as the "Start golfing" banner disappears.
When scrolling down after revealing the languages, the logos go over the top banner, not under.
The Friends/Information/Activity tabs on the user page overlap with the top of the information pages they show.
It's very unintuitive how to actually get started with any challenges; the "Start golfing" button takes me to my user page, but I can't see any obvious way to actual, well, start golfing. Took me a bit of time to realise that I needed to hover over the cursor sy…

Seggan

newer langs as in newer vesions of vyxal, fig, flax, rsn, etc

ah i see what ure saying

nvm

Adám

12:53 AM

@Zionmyceliaadamancy The cursor symbol? The only way I found was via the "bullet list" icon.

Zion mycelia adamancy

@Zionmyceliaadamancy Also, for anyone who is thinking about making any kind of multi-language coding site, I strongly advise you read my reasoning here for freedom of language use. In short, it's what makes CGCC the best place on the internet for creating, improving and developing languages

TKirishima

@Zionmyceliaadamancy I think I totally understand your point of view. I will try to add language each month via votes. However, some languages have quite strange configurations (or at least I'm not familiar with them) and I try to have enough languages to be sure I know what I'm doing. I know some sites like TIO manage to have a lot of languages but they are not necessarily kept up to date like Python for example (which is at version 3.7.8).

I prefer to have few languages but they are updated. However; I understand your point.

Zion mycelia adamancy

@Adám Well, that's not great

@TKirishima FWIW, TIO is only out of date because Dennis (the sole maintainer) went offline over 2 years ago

For 6 years before that, languages were regularly added and updated

TKirishima

@Zionmyceliaadamancy Thank's ! I'll fix them as fast as I can!

Zion mycelia adamancy

@TKirishima Also, TIO is fully open source, including the code it uses to invoke each language. You can definitely take a look at that code to figure out how to run some of the more obscure TIO langs

TKirishima

12:57 AM

@Zionmyceliaadamancy I'll try to make it more intuitive

@Zionmyceliaadamancy Concerning the "The Friends/Information/Activity tabs on the user page overlap with the top of the information pages they show." can you show an image please ?

Zion mycelia adamancy

1:13 AM

@TKirishima I will in the morning, no longer at my computer

TKirishima

Ok! np! Thank's for the help!

NoHaxJustRadvylf

1:52 AM

The year is 2042. Computer science education has become considerably more common, and most young adults are at least capable of writing simple Python scripts. Programming is a hobby comparable in size to video games, and like speedrunning did in 2020, code golf, against all odds, has suddenly become incredibly popular. As one of the first organized and extensively documented places in which code golfing took place, our site, which is now archived and defunct,

7

is now moderately famous, and those of us who return to the world of code golf are looked upon as gods. Although the code golf of the '40s looks much different from today's, and some of us reject it as, now mainstream, it has changed into something close to unrecognizable. But others see it as a chance to finally talk with a large community about their hobby, and stories of golf wars and debates in TNB are as war stories from experienced veterans.

New golfing languages, many not particularly good, appear every day, and the biggest site for golfing, Code Golf Contests, is a proving ground for ever evolving strategies both in the usage and the design of novel golflangs. As quantum computing has become more practical, rules are hashed out for its incorporation into CGC, and the qubit becomes the second recognized scoring method.

Events like language golf challenges, mini golfs, and KotHs receive widespread attention not only on text-based social media platforms like the outdated Reddit and its modern competitors, but also on video sharing platforms, regularly gaining millions of views. Parents worry the obsession with other people's scores is causing children to be self conscious of their own code, but this is dismissed by many as a moral panic.

Global code quality suffers, hard drive prices plummet, but by the late '40s, the excitement has died down, and the programming world moves onto the next trend, retro web development, and #savejavascript goes viral as Wasm is set to ultimately discontinue the aging relic.

emanresu A

Wow

lyxal

Dang it where was this code golf fanfic when I needed it at several times in the last 2 years?

emanresu A

A cult has sprung up around the one known as "Lyxal", and unlike Radvylf, I don't have any creative writing skill.

NoHaxJustRadvylf

@TKirishima Wow this looks really cool!

TKirishima

@NoHaxJustRadvylf Thank's a lot @NoHaxJustRadvylf :D

NoHaxJustRadvylf

2:02 AM

The UI feels a bit unintuitive, but it looks good, and that's the hard part out of the way

TKirishima

I'll try to upgrade it !
But for the moment I don't know how !

I'm going to sleep!
Goodbye everyone!

NoHaxJustRadvylf

'night! o/

Leaky Nun

2:16 AM

CMC: given a list of positive integers [a,b,c,...] return a+2b+3c+...; e.g. [3,1,4,1,5] -> 1*3+2*1+3*4+4*1+5*5 = 3+2+12+4+25 = 46

Steffan

@LeakyNun Vyxal, 3 bytes: Try it online!

It would be 2 bytes in 05AB1E

ƶO

Python, 43 bytes: lambda x:sum(x*-~i for i,x in enumerate(x))

Leaky Nun

2:45 AM

@Steffan I have a 36-byter

Steffan

Yeah I'm really bad at Python golfing

38

user

Js ‘s reduce should make this easy

Maybe a=>a.reduce((b,c,i)=>b+c*i)

I wonder how bad a recursive solution would be in Haskell

If you took the list reversed

Steffan

3:30 AM

a=>a.reduce((b,c,i)=>b+c*-~i) probably

TIO

Lemme make a horrible Prolog one

65

And the same thing in Haskell, 43.

lyxal

@Steffan mfer using prolog like what about us smooth brains who have to use nooblog?

Steffan

you'll just have to do without the pro functionality *walks off*

Aiden Chow

Desmos, 27 bytes f(l)=total([1...l.length]l)

lyxal

@Steffan can be 2 with a flag

Steffan

I know

Aiden Chow

3:45 AM

Geogebra, 43 bytes (input as comma separated integers in the input box):

l={
InputBox(l
Sum(Zip(al[a],a,1...Length(l

Steffan

my haskell one can be f s=sum$zipWith(*)s[1..length s] (32) without recursion

anyway I should probably be in bed

Aiden Chow

wait bruh what am i thinking, geogebra also has vectorized operations

l={
InputBox(l
Sum((1...Length(l))l

Bubbler

4:33 AM

@LeakyNun Jelly, 2

Python, 36

Haskell, cheating 16 and legit 19

@LeakyNun actually Haskell, 13

Jo King

5:08 AM

@Steffan pro-erlog, 43 bytes

for some reason, prolog has scanl but not scanr??

New Posts

5:27 AM

0

Q: Print all of this: Vrei sa pleci dar nu ma, nu ma iei

Make a program that prints the lyrics of [link] exactly 1234567890 times. You know the rules, and so do I: The text must appear exactly like in the pastebin (exactly) No external resources, all must be generated by code. No existing compression algorithms. (gzip, 05AB1E compressed strings (they ...

lyxal

5:56 AM

Hehe dupe hammer goes brrr

Aiden Chow

lolol nice

New Posts

0

Q: Do math without type changing (completely)

Write a program that gets two whitespace-seperated floating point numbers from input, then prints their sum, difference, product, quotient (to at least 3 decimal places) and modulo WITH ONLY STRING OPERATIONS. Divison or modulo by zero will print e (for error) That means no input a; input b, prin...

code-golf math

Aiden Chow

damn another one

Aiden Chow

6:24 AM

whats the close vote for this one anyways

mousetail

6:39 AM

@NoHaxJustRadvylf There are already a ton of sites specifically about KOTHs, probably the format most likely to have mainstream appeal.

lyxal

7:30 AM

Me: signs up for weekGolf

Me: tries to sign in

Username doesn't exist

Me: tries to sign up again

Username already taken

Yes 👍😎

Also me: tries sign in with github

Site can't handle request

Finally got in with SE authentication

Also, Vyxal does have a logo for the language suggestion page

mousetail

What is weekGolf?

Aiden Chow

@mousetail a website some dude advertised here

mousetail

Seems like a site that fits me since I'm weak at golf

lyxal

Cc @TKirishima to the above

Adám

Wonder if they'll be able to come up with a good challenge every week for the foreseeable future. Sounds tough.

Aiden Chow

7:44 AM

@Adám maybe they got some planned for the first few weeks, so it wouldnt immediately flop

but ye that sounds tough, i dont think i would even be able to think of of a good challenge idea every month lol

Adám

We are 25 employed people at Dyalog, and we have a hard time coming up with 15–20 challenges per year.

pxeger

8:28 AM

@Bubbler Python, 34 bytes: f=lambda x:x>[]and sum(x)+f(x[1:])

Adám

@LeakyNun APL, 7 bytes: ⍳∘≢+.×⊢

pxeger

@TKirishima On the profile page, "Disconnect" is presumably the literal translation of French déconnecter, but the correct English in this context would be log out or sign out

Adám

@LeakyNun APL, 6 bytes: +/+\⌽⎕

BQN 5 bytes: +´∾∘↓

New Posts

0

Q: Convert from Greeklish to modern Greek

Greeklish, a portmanteau of the words Greek and English, is a way of writing modern Greek using only ASCII characters. This informal way of writing was extensively used in older applications / web forums that did not support Unicode, and were not programmed to show Greek characters. Nowadays, its...

code-golf string substitution

mathcat

9:25 AM

2 days ago, by mathcat

I'll post this tomorrow, but in the meanwhile, feedback is appreciated.

i guess today is finally tomorrow

New Posts

9:45 AM

0

Q: Squash it ... again!

If you place the positive integers together and read each set of two adjacent digits at the same time, you get: (A136414) 12, 23, 34, 45, 56, 67, 78, 89, 91, 10, 1, 11, ... However, if you squash that sequence again: 12, 22, 23, 33, 34, 44, 45, 55, 56, 66, 67, 77, 78, 88, 89, 99, 91, 11, 10, 1, ...

code-golf math sequence

mousetail

11:00 AM

F, I got a nice solution for the squashed series problem but I read the question entirely wrong so it's useless

mathcat

what's it?

mousetail

My solution combined the first and last character of each number instead of all of them

It's so elegant though let f=|s:Box<dyn Iterator<Item=u32>>|Box::new(s.scan(0,|a,b|Some((*a*10+b%10,*a=b/10_u32.pow(b.log10())).0)).skip(1));

Leaky Nun

11:19 AM

@Bubbler yay you found my python solution

@pxeger oh wow you found one even smaller

TKirishima

@Adám Added

@pxeger Added

@lyxal Well that's kind of strange :/ I don't know why this is happening. I'll investigate

@AidenChow Yes

Adám

@TKirishima Except by making all three parts white, it becomes a lone X, instead of being overlaid ᵛᎩX for Vyxal.

mousetail

@TKirishima The stripping of leading white space from errors is a bit annoying. Makes rust errors impossible to read

TKirishima

@Adám Yes. And some logos like Raku aren't in color too, because it is the point of putting them in white

Adám

I get it it, but at least with Raku, the logo is very clear. You kind of need three levels for Vyxal's logo. Not a big deal, for a temporary thing, though. FWIW, a lot of work went into making APL's logo such that it could work when made monochrome.

You could use three shades of white/lightgrey for Vyxal (your Scala icon has this), or use outlines.

TKirishima

11:34 AM

@mousetail I'll try to fix it

@Adám OK!

emanresu A

@Bubbler 28 in whyyhon

Hello, World!

1:20 PM

0

A: "Hello, World!"

Stack Cats, 94 + 3 = 97 bytes (^_^_[>_[_*:>^]<^[>+:^[_!:+_+:>-!^_!]<:_^_I_+>_]<:_I_+:>_I^+>_I_+*-!*>+![+]+_!_-^+_)*_:_-^_!_: Needs to be run with the -m switch, so +3 Try it online! I'm sure there're still a lot of bytes which can be golfed, but this is the best I could think of so far. Due to t...

pxeger

2:14 PM

CMC: most amusing deadlock (busy loops are not allowed)

Seggan

final a = Object();
final b = Object();

// Thread 1
synchronized(a) {
synchronized(b) {
}}
// Thread 2
synchronized(b) {
synchronized(a) {
}}

Java ^

pxeger

congratulations on misspelling the word "synchronized" 4 different ways

(four, because you spelt them all with an s, but Java wants a z)

mousetail

Isn't there just 1 type of non-trivial deadlock?

pxeger

How do you define a type of deadlock?

user

@pxeger When you're in the grocery store standing next to the apples, waiting for the guy buying oranges to leave so you can get your mandarins, but the guy standing at the oranges is just standing there waiting for you to leave so he can buy his honeycrisps

Seggan

2:19 PM

winner ^

mousetail

Only one is A waits for a lock held by B while B is waiting for a lock held by A

user

@Seggan (obviously not original, I saw it in SMBC or somewhere else)

mousetail

Most kinds of deadlock are a extension of this, perhaps with more intermediate steps

pxeger

@mousetail I suppose ultimately yes, but what makes a deadlock amusing are the intermediate steps

mousetail

ok fair enough

pxeger

2:21 PM

Do you consider pause(2) to be a deadlock?

mousetail

Would priority starvation be considered a deadlock? I guess that would count as a distinct type

pxeger

maybe

but then what makes priority starvation different to exit?

mousetail

A process would run but never gets a turn due to multiple higher priority processes taking turns

assuming a scheduler without starvation prevention

pxeger

I guess the difference is that exit means the process's memory etc stop existing

whereas a deadlock means the process still exists, but never runs

mousetail

that seems like a reasonable definition

pxeger

2:28 PM

the difference between starvation and most other deadlocks though would be that starvation can happen due to a preemption but all others can only happen through system calls that never return

mousetail

true

Unless the lock causing a deadlock is a spinlock

pxeger

that would count as a busy loop though

mousetail

True

New Posts

0

Q: Unique rainbow programs

Given a list of positive integers as input you are to determine if there is a way to pair up the elements such that: All but one element is in a pair No element is in more than one pair Now two pairs have elements separated by the same distance in the original list For example if the input is: ...

code-golf decision-problem self-validating

pxeger

@NewPosts @WheatWizard oops, stomped your edit

Wheat Wizard

2:37 PM

Ah I was just coming to ask what the point of that edit was. I see.

pxeger

the "An edit has been made to this post" thing came right as I clicked save

Mayube

Happy timezone folks

Seggan

3:24 PM

@LeakyNun Fig, theoretical (you can't actually run this yet) ~3.292 bytes S*]r

Sum the vectorised * multiplication of the input and the range [0,n) ] incremented

]r actually pops up quite a lot in my test programs, maybe i should stick that in a monograph somewhere

Mayube

3.292 bytes??

what kind of encoding does Fig use?

Seggan

ascii irrational

the actual score is 4log_256(96)

how many bits

~6.585

nonsensical

3:32 PM

that many per byte

thejonymyster

i meant total :P

Seggan

multiply that number by 4

thejonymyster

26.34 bits?

Seggan

yep

thejonymyster

i draw the line at fractional bits

Seggan

3:35 PM

it is theoretically possible to build a filesystem that does that

Jun 11 at 22:30, by Radvylf Programs

Yeah. It's actually a relatively recent rule change, prior to that you had to pad to full bytes in most cases.

thejonymyster

can you not just go arbitrarily small

NoHaxJustRadvylf

@thejonymyster Consider the old Soviet ternary computers, those would be irrational numbers of bits

thejonymyster

fractional bytes i understand

NoHaxJustRadvylf

@thejonymyster Sort of, but that's very cheaty and there's no phyical way to store it, so it'd be hard to justify

thejonymyster

idk it just seems like there needs to be a smallness limit or else its not clear where the "cheaty/noncheaty" boundary really is

anything less than a bit feels cheaty to me since its arguing for an entirely different computer than even just a fractional byte

that software vs hardware gap isnt one that makes sense to cross but thats my onion

~~plus its confusing to my unevolved brain~~

NoHaxJustRadvylf

3:40 PM

@thejonymyster Well, anything more granular than a bit isn't physically possible, so it falls under the "pretty cheaty and hard to justify"

Seggan

so now yall saying Fig is bad?

or do you mean less than a bit per byte

thejonymyster

im saying that the full code of your solution needs to have a nonnegative integer number of bits

NoHaxJustRadvylf

I mean less than a bit per character, yeah

@thejonymyster No

That's a bad rule IMO

thejonymyster

then i dont understand :P

@NoHaxJustRadvylf what did you mean by this then

NoHaxJustRadvylf

Well, how would you score, in bits, a language that uses trits?

thejonymyster

3:43 PM

i'm saying we shouldnt use trits

NoHaxJustRadvylf

@thejonymyster As in, if every command in my language is half a bit

thejonymyster

because that is a hardware leap

NoHaxJustRadvylf

There is hardware that supports it though

thejonymyster

right but is there not a way to take it even smaller?

NoHaxJustRadvylf

And you could physically design hardware to support any log_256(n) bits per character

@thejonymyster No

thejonymyster

3:44 PM

oh wait so half bits is the limit?

NoHaxJustRadvylf

Beyond 1 bit per character, it loses meaning.

thejonymyster

ok so i dont understand... if its 1 bit per character, and theres no way to get a fractional character count, how does that not equate to an integer number of bits

NoHaxJustRadvylf

You can have more than 1 bit per character that's still irrational

Seggan

if its 1 bit per character you only end up with 2 characters

NoHaxJustRadvylf

Like approx. 1.58 bits per character for trits

But you can't have, e.g., 0.001 bits per character, or 1.01 bits per character, or 0.9999 bits per character, because there's no reasonable physical representation of that

thejonymyster

3:47 PM

whats the dif between the 1.58 case and the 1.01 case?

NoHaxJustRadvylf

(Well you technically could but nobody would accept it as not cheaty)

thejonymyster

im starting to get it

Mayube

@thejonymyster the former has a real-world precedent

NoHaxJustRadvylf

@thejonymyster 1.58(with more stuff after the decimal) is log_2(3)

It's how many bits of information you store in a trit

thejonymyster

so are we saying trits is the limit because its implemented?

NoHaxJustRadvylf

3:48 PM

It's not actually 1.58 bits on any machine, but it equates to that much information

@thejonymyster No, you could have log_2(5), log_2(6), log_2(97), log_2(255), etc.

thejonymyster

can that number keep going up then? the 255 one?

NoHaxJustRadvylf

Yep, you could have log_2(256) (or 8 bits per character, what most languages use), or log_2(65536), or log_2(174673282748738), or any other integer

thejonymyster

oh right im going the wrong way lol

NoHaxJustRadvylf

It's basically a generalization of the idea that to represent 64 options, you need 6 bits (log_2(64))

thejonymyster

right i just misremrmbered the argument order for logs sort of

i was all backways n shiz

Seggan

3:50 PM

tl;dr bits are a measure of information, not necessarily a hardware thing

thejonymyster

right ok

so log_2(3) is the most granular it can be because log_2(2) is 1 fgdjg

ok im appeased

danke all

NoHaxJustRadvylf

Technically our rules don't disallow things like 0.001 bits per character, but it's like the flags and loopholes where it's mainly enforced via downvotes

mousetail

4:45 PM

You can't technically get any less than 0.128 bytes per character though right?

Mayube

@mousetail that would be less than 1 bit per character, which we already established isn't a thing

NoHaxJustRadvylf

Yeah, below 1 bit per character individual characters become meaningless

(although, you could argue a language where every two characters represent a single bit is half a bit per character, though it's a bit different)

mousetail

There's plenty of langues with very limited information density but I don't think they should get special treatment for it

NoHaxJustRadvylf

Well no, most would continue to be 1 byte per character unless a different method of counting bytes is explicitly included, but that's not for "special treatment" reasons, just for convenience

mousetail

If your language is so bad it required multiple bytes to represent a single bit you just build a more dense language that compiles to it

NoHaxJustRadvylf

4:51 PM

Since languages are scored independently it doesn't really matter if you apply a constant scaling factor to a single language's byte count, since whenever people do want to unofficially compare languages, they can undo that scaling in their heads to make it fairer if they feel that's necessary

@mousetail That assumes that short code is the objective for more than a tiny and already incredibly well optimized subclass of languages

Just having a weird or deliberately wasteful method of counting bytes could be the goal in some cases

mousetail

hmmm

pxeger

Jun 27 at 17:17, by DLosc

I dunno... anytime a question about scoring comes up, I keep coming back to, "Are we comparing scores across different languages or not?" And the answer always seems to be "No, definitely not. But also yes."

NoHaxJustRadvylf

Jun 27 at 17:17, by Radvylf Programs

We officially don't, but everyone does unofficially because half of us are golfing language designers

Our current rules work great. They allow basically every language to compete with itself and similar languages fairly and objectively. They also allow the subjective part, comparing across different languages, to stay subjective.

That's a feature, not a bug.

Wheat Wizard

5:27 PM

I have to figure out how to add flags to hgl.

Jokes aside what I actually need for hgl is to get another human being to use it other than myself.

Mayube

@WheatWizard you called?

Wheat Wizard

You want to use hgl?

Mayube

why not?

how esoteric is it?

Wheat Wizard

5:42 PM

Uh It's not all that esoteric.

It's esoteric in some ways.

Hard to say exactly.

It's not a programming language on it's own it's a library for Haskell.

pxeger

It appears quite well documented

I just have no idea where to start when using it

Mayube

@WheatWizard oh boy, guess it's time to learn Haskell then :P

Wheat Wizard

@pxeger The best way I think it to try to write a haskell answer completely in point free. Because most useful haskell functions have a hgl version.

Also if you put it into a prompt you should get a warning for each function used telling you its hgl name.

(results may vary)

If you do find a function that doesn't exist, either in long form or short form. You can open an issue, it's a very easy fix for me. (and sometimes the short version does exist it's just got a really weird name)

user

Is there a way to use it on online interpreter?

pxeger

it's on ATO

user

5:49 PM

Oh nice

Wheat Wizard

You are probably going to want the docs though.

user

> Please do not use this for actual code.

I don't think I'm going to bother following these docs :P

@WheatWizard I can't find a docs folder or wiki here, are the docs elsewhere?

Wheat Wizard

You have to build them right now

user

Oh ok

Code doesn't look all that terrible to read

Wheat Wizard

There should be instructions in the readme.

Yeah the code can also function as docs

I should write a basic intro somewhere

NoHaxJustRadvylf

5:57 PM

@WheatWizard The caird coinheringaahing approach to documentation :p

(Probably works better with Haskell tho)

Wheat Wizard

There is pretty extensive docstings is what I meant. The code itself is not great documentarion

Everything int the docs is somewhere in the source

The code itself is mostly like ghf = ghi < bk << tg

Sandbox Posts

7:39 PM

0

A: Sandbox for Proposed Challenges

Next digit of rational number codegolf string Story: The π was recently computed with accuracy to 100 trillions digits, but it is useless to us. We can't do accurate enough math, because rational numbers are too boring and so we don't know that much digits of them! Challenge: Your task will be to...

Seggan

Oh wow. look at the byte counts

Sandbox Posts

7:59 PM

0

A: Sandbox for Proposed Challenges

Every possible pairing code-golf combinatorics Given an positive even integer \$ n \$, output the set of "ways to pair up" the set \$ [1, n] \$. For example, with \$ n = 4 \$, we can pair up the set \$ \{1, 2, 3, 4\} \$ in these ways: \$ \{\{1, 2\}, \{3, 4\}\} \$ \$ \{\{1, 3\}, \{2, 4\}\} \$ \$ ...